α-RgIB: A Novel Antagonist Peptide of Neuronal Acetylcholine Receptor Isolated from Conus regius Venom

Conus venoms are rich sources of biologically active peptides that act specifically on ionic channels and metabotropic receptors present at the neuromuscular junction, efficiently paralyzing the prey. Each species of Conus may have 50 to 200 uncharacterized bioactive peptides with pharmacological interest. Conus regius is a vermivorous species that inhabits Northeastern Brazilian tropical waters. In this work, we characterized one peptide with activity on neuronal acetylcholine receptor (nAChR). Crude venom was purified by reverse-phase HPLC and selected fractions were screened and sequenced by mass spectrometry, MALDI-ToF, and ESI-Q-ToF, respectively. A new peptide was identified, bearing two disulfide bridges. The novel 2,701 Da peptide belongs to the cysteine framework I, corresponding to the cysteine pattern CC-C-C. The biological activity of the purified peptide was tested by intracranial injection in mice, and it was observed that high concentrations induced hyperactivity in the animals, whereas lower doses caused breathing difficulty. The activity of this peptide was assayed in patch-clamp experiments, on nAChR-rich cells, in whole-cell configuration. The peptide blocked slow rise-time neuronal receptors, probably α3β4 and/or α3β4α5 subtype. According to the nomenclature, the new peptide was designated as α-RgIB.


Introduction
Marine mollusks from Conus genus may produce from 50 up to 200 biologically active molecules that can be injected in the prey to capture or be employed as defense and/or escape mechanisms to deter competitors. The peptide toxins, called conopeptides, are composed of 10-40 amino acids (including nonnatural amino acids) and are abundant in the venom. Peptides presenting a rigid structure due to more than one disulfide bridges are common, being called conotoxins. These peptides act specifically on ionic channels and/or neuromuscular receptors [1,2].
Conotoxins are classified according to three schemes: the similarities between the endoplasmatic reticulum signal sequence of the conotoxin precursors (gene superfamilies), the cysteine patterns of conotoxin mature peptide regions (cysteine frameworks), and the specificities to pharmacological targets (pharmacological families) [3,4].
Conopeptides of the pharmacological family , which acts on neuronal acetylcholine receptor, have been found in the A, D, L, M, and S gene superfamilies [5,6].
Typically, -conotoxins are peptides with 12 to 16 amino acid residues and two disulfide bridges, presenting the pattern CC-C-C. These peptides are competitive antagonists of the nicotinic acetylcholine receptors (nAChR) and display high selectivity by subtypes of this receptor [5,[7][8][9]. After the blockage of the muscular acetylcholine receptor, the -conotoxins significantly decrease the amplitude of the motor end 2 International Journal of Peptides plate postsynaptic potentials in vertebrates, paralyzing the prey [10].
In the Brazilian tropical coast, there are approximately 18 species of cone snails [11]. Conus regius (Gmelin, 1791) is a vermivorous species that inhabits rock and coral deep waters of Florida (USA), Central America, and the Northeast and East coast of Brazil, including Fernando de Noronha archipelago [12].
In this work we described a novel peptide from Conus regius venom, belonging to the -conotoxins family. This peptide blocks the neuronal acetylcholine receptors on PC12 cells, which comprise 3 4 and/or 3 4 5 subtypes receptors, probably target of the peptide. , and the process number was 02001, 000775/00-00. Venom was extracted from the specimens as previously described [13]. The crude venom was obtained by dissection of the venom duct gland and then freeze-dried and stored at −80 ∘ C.

Material and Methods
Voucher material is deposited in the malacological collection of Zoology Museum of University of São Paulo, São Paulo, Brazil.
For -RgIB purification, the interest peak was fractionated in a Merck C18 column (300 × 4.6 mm), in a 19 to 21% B gradient over 20 min, at a constant flow rate of 8 mL⋅min −1 . A subsequent purification step was still necessary to obtain the peptide. This purification was conducted in a Merck C18 column (300 × 4.6 mm), in an isocratic elution at 35% B (TFA/methanol/H 2 O 1 : 900 : 100) at a constant flow rate of 1 mL⋅min −1 . 2.5. "De Novo" Peptide Sequencing. Mass spectrometric "de novo" peptide sequencing was carried out in positive ionization mode on a Q-TOF Ultima API fitted with an electrospray ion source (Micromass, Manchester, UK). Briefly, the amounts of previously lyophilized peptide were dissolved in 50 mM ammonium acetate, reduced with 50 mM DTT, alkylated by 150 mM iodoacetamide, and hydrolyzed by 25 nM trypsin, according to slight modifications of Westermeier and Naven [14]. The reaction products were then lyophilized and dissolved in 50% ACN, containing 0.1% FA and injected into the source at 5 L⋅min −1 by a Hamilton infusion pump, or directly injected using a Rheodyne 7010 sample loop coupled to a LC-10A VP Shimadzu pump operating at 20 L⋅min −1 constant flow rate. The instrument control and data acquisition were conducted by MassLynx 4.0 data system (Micromass, Manchester, UK) and experiments were performed by scanning a mass-to-charge ratio (m/z) of 50-1800 using a scan time of 2 s applied during the whole chromatographic process. The mass spectra corresponding to each signal from the total ion current (TIC) chromatogram were averaged, allowing an accurate molecular mass determination. External calibration of the mass scale was performed with NaI. For the MS/MS analysis, collision energy ranged from 18 to 45 and the precursor ions were selected under a 1-m/z window.

Mass Spectrometry
2.6. Biological Activity. The biological activity of -RgIB was determined in Swiss Webster mice (5.5 to 7 g body weight) by observation of the behavioral disorders after intracranial injection [15] of the peptide diluted in NaCl 0.9%, in concentration of 0.1, 0.5 and 1 nmol. Alterations were compared to animals injected with NaCl 0.9% (control). All animals were observed by 60 min.

Patch
Clamp. BC 3 H1 cells, mouse myocytes which express nicotinic acetylcholine receptors, were acquired by ATCC (CRL-1443) and maintained in culture according to Sine and Taylor [16] to electrophysiological experiments. In order to verify the subtype of neuronal nicotinic receptor that the peptide acts, PC12 cells were employed and maintained in culture according to Greene et al. [17].
Throughout the experiment, the membrane potential was clamped at a −60 mV for BC 3 H 1 cells and −70 mV for PC12 cells, holding potential using an Axon Axopatch amplifier (Molecular Devices, California, USA). Data were recorded and digitized by Clampex 8.2 software (Molecular Devices, California, USA) and plots were made using Origin 7.0 software (OriginLab Corp., Northampton, MA).
Control currents were performed with 1.5 mM carbamylcholine, using the cell-flow technique [20]. After the carbamylcholine administration, the peptide (10 M) was incubated on cells, and then another dose of the agonist was incubated [21].

Data Fitting, Statistical Analyses, and Sequence Alignment.
When data fitting was performed, results were presented as the calculated value ± standard deviation (SD). Otherwise, data correspond to the mean of three individual experiments. Peptide sequence alignment was performed using ClustalW software [22]. The 3D model of the peptide -RgIB, as well as three PDB deposited 3D solution structures, was created by I-TASSER [23,24].

3.1.
Purification. The crude venom from C. regius was fractionated by RP-HPLC, as shown in Figure 1(a). Some peaks could be detected along the profile, and the arrow indicates the peak of interest. Two subsequent chromatographic steps were necessary to purify the peptide (Figures 1(b) and 1(c)), under the conditions described material and methods section. After the third step of purification, the purity and the molecular mass of the peptide were assessed by MALDI-TOF/MS (Figure 1(d)).

"De
Novo" Peptide Sequencing. After cysteine bridge reduction and alkylation, the reaction product was digested with trypsin. The obtained peptides were submitted to MS/MS analyses ( Figure 2) and ions were selected and fragmented by collision with argon (CIF), yielding daughter ion spectra ( Figure 3) that was processed with BioLynx and manually checked for accuracy of interpretation. Since the digestion allowed peptides with missed cleavage sites, it was possible to assemble the fragments without the aid of another digestion with a different enzyme. The sequenced peptides, their charge states, and theoretical molecular mass are presented in Table 1.
The peptide sequence was determined to be TWEECCKNPGCRNNHVDRCRGQV. This sequence has 4 cysteine residues with pattern CC-C-C, typical from conotoxins of framework I [25]. This peptide was named -RgIB, according to the guidelines for conotoxins nomenclature ConoServer and has been assigned the following UNIPROT accession number: C0HJA8 [3, 6, 26]. Table 1 of the supplementary material available online at http://dx.doi.org/10.1155/2013/543028). Based on this large alignment, a phylogeny was constructed (supplementary Figure 1) and the peptide sequences present at the branch containing -RgIB were realigned (Figure 4). ClustalW standard annotations consider, as expected, the Cys residues as consensus ( * ), and the Glu at the 8th aligned position, as being highly conserved (:). Moreover, according to the algorithm standard notation, the Pro residue at the 13th aligned position is also conserved (.). Among the UNIPROT database, the SwissModel tool could not identify any suitable template for structure prediction of -RgIB, therefore an external application was used. Figure 5 presents a 3D model of -RgIB; created by I-TASSER [23,24], as well as three PDB deposited 3D solution structures of -RgIA (P0C1D0) mutants, a conotoxin that specifically and potently blocks the 9 10 nAChR [27]. In spite of -RgIB N-and C-terminal extensions and longer interbridge peptide sequence, the model and the structures are tridimensionally related, for example, a C-shaped structure, held by the Cysbridges.

Biological Activity.
The in vivo biological activity of the peptide was assessed by means of intracranial injection in Swiss Webster mice. Following 1 nmol injection, the animals displayed a hyperactive behavior, defecating and urinating all the time, which was not observed for the control group that received saline solution. Auditory stimuli, for example, a hand-clap or hitting the cage, also triggered the hyperactive behavior. Interestingly, the lower doses (0.1 and 0.5 nmol), caused the animals to have difficulty in breathing. Although the peptide promoted behavioral disorders, it was not lethal to the animals.

Patch
Clamp. Whole-cell voltage clamp measurement was used to verify the ion currents on acetylcholine receptors. BC 3 H1 cells, which express the acetylcholine muscle type receptors on the surface, and PC12, which terminally differentiate in neurons and express nicotinic neuronal receptors [21] were selected for the experiments. Carbamylcholine, a stable and well-characterized analogue of acetylcholine, was used as an agonist [28], for it elicits a fast activating current that rapidly desensitizes during the application.
10 M -RgIB was not able to induce any change in the ion currents on BC 3 H 1 cells (data not shown), as well as a higher dose (30 M) of the peptide. d-tubocurarine (a classic nicotinic receptor antagonist) was used as a positive control and successfully to block this channel (data not shown).
After the incubation of the peptide with PC12 cells, fast and slow desensitization of the receptor was observed. Figure 6(b) shows that on neuronal slow rise-time receptors, -RgIB is able to block the ion current by 40%, compared to cells stimulated with carbamylcholine ( Figure 6(a)). The blockage was irreversible and persistent, once the current does not recover after a new application of the agonist, carbamylcholine (Figure 6(c)).

Discussion
Conotoxins are classified according to the similarities between the signal sequence of the conotoxin precursors (gene superfamilies), the cysteine patterns of conotoxin mature peptide regions (cysteine frameworks), and the specificities to pharmacological targets [3,4]. This new peptide was termed -Rg-IB because the peptide acts on neuronal acetylcholine receptors (" "), was extracted from a Conus regius specimen ("Rg"), displays a cysteine framework I-CC-C-C ("I"), and was the second International Journal of Peptides    peptide discovered with both being from C. regius with a cysteine framework I ("B") [25].
-RgIA was the first -conotoxin described from C. regius, acting on neuronal nicotinic receptors. This peptide has been thoroughly characterized in terms of its primary and three-dimensional structures [29], as well as regarding its biological effect, for example, the blockage of the 9 10 nAChR [30,31]. -RgIA and -RgIB come from the same animal, belong to the same toxin family, and possess similar biological effects; however, their amino acid sequences differ. Figure 4 shows the ClustalW alignment of -RgIB and its closest phylogenetic relatives (supplementary material), besides -RgIA, which was not considered to be similar (according to MEGA5), but was manually inserted in the figure for the benefit of sequence comparison. -RgIA is shorter, both in the N-and C-terminal flanking regions, as well as in the inter-Cys-bridge region. Nevertheless, in a considerably small universe of possibilities (8 out 12, since 4 amino acids are necessarily Cys), -RgIA and -RgIB bare considerable similarities: the Pro, at the 13th aligned position, and the charged residues at the 17th and 18th aligned positions. It is noteworthy to mention that,  in spite of the phylogenetic analyses, all conotoxins listed in Figure 4 (except -RgIA and -RgIB) come from other Conus species: C. leopardus (A6 M938, A1X8C2, A1X8C3), C. litteratus (Q2I2R6), C. pulicarius (P0C8U6, P0C8U7, P0C8U8, P0C8U9, and P0C8V0) and C. geographus (P01519). Moreover, only P01519 has been detected at the protein level and has been characterized as active on the muscular nicotinic receptors [32].
There is no high level of homology between -RgIB and the conopeptides described until the present moment; therefore, the identification of a proper 3D structure to serve as a template for homology modeling is deprecated. Instead, a structure was predicted by using I-TASSER server [23,24]. Figure 5 shows that, in spite of the low homology with -RgIA, -RgIB model assumed the same basic shape as the NMR determined structures of the -RgIA mutants, available at the PDB database [27]. Regarding the rather unique amino acid sequence of -RgIB and thoroughly analyzing our data, we could not rule out the possibility that one of the glutamic acid (Glu) residues of this novel conotoxin would be a gamma-carboxyglutamic acid residue (Gla). Our suspicions arouse from the slightly higher deviation between the theoretical and calculated molecular mass values for A and B ions (Table 1), that could reflect that a side chain carboxylation and not an Nterminal acetylation would be present. Moreover, conotoxins are known for presenting posttranslation modifications, Gla included [35][36][37][38] and, even though the MALDI data of the crude peptide support the proposed peptide sequence, MALDI ionization is also a source of facile decarboxylation for Gla residues [39]. Our future experiments with C. regius conotoxins ( -RgIB included) will clarify this matter.
Neuronal nicotinic acetylcholine receptors (nAChRs) belong to the pentameric superfamily of Cys-loop ligand gated ionic channels. They are composed of either homomeric or heteromeric and subunits assembled from a family of 12 distinct neuronal nicotinic subunits ( 2-10; 2-4) [5]. The combination of subunits 2, 3, and 4 with 2 and 4 results in a functional receptor, as well 7, 8, and 9 homomeric receptors [40,41]. In our experiments, it was verified by RT-PCR (supplemental Figure 3) that the pool of PC12 cells expressed 3, 5, 7, 2, and 4 subunits of neuronal nicotinic receptors, the same pattern found by Sargent [40] in PC12 cells. However, in spite of -Rg-IB affinity by the PC12 nicotinic receptors, there are still other neuronal nicotinic receptors that may be higher affinity targets for these toxins that were not explored in the present work.
AuIB, from Conus aulicus, which is also an -conotoxin, blocks the 3 4 receptors; however, the currents can be recovered after the toxin washing [42]. In our experiments, the no-recovery of -RgIB is probably due to the irreversible action of the peptide on the receptor. Successive applications of the agonist (carbamolycholine), in control experiments, did not cause recovery of the ion currents on slow rise-time receptors (data not shown). Besides the irreversible action, the peptide may also be able to prolong the desensitization time of the receptor, since the repeated CBC administration on -RgIB-treated PC12 cells was not able to recover the initial current, which is either caused by the irreversible binding of a low affinity toxin or the prolonging of the desensitization time of the receptor (or both).
Sudweeks and Yakel [43] showed that 3, 7, and 2 subunits of nAChR are correlated to fast rise-time receptors. The slow desensitization is a characteristic of 3 4 receptor, while 3 2 receptor is from fast desensitization [44]. The fast desensitization receptors, on PC12 cells, contain 3 2, 3 2 5, and 7 subunits, while slow desensitization receptors are formed by subunits 3 4 and 3 4 5. -RgIB was able to inhibit the currents elicited by carbamolycholine on PC12 cells, mainly on the slow desensitization component, which comprise, in our model, 3 4 and 3 4 5 receptors.
The intracranial injection assay was performed to investigate whether there would be any direct activity of the toxin in the central nervous system (CNS), once peptides can promote behavioral alterations by acting on receptors and ionic channels on CNS. These alterations can indicate activities on specific ionic channels. For example, -conotoxin GVIA causes trembling on the mice, which indicates an action on calcium ionic channels [45]. The -nicotinic acetylcholine receptor (nAChR) is associated to attentiondeficit/hyperactivity disorder [46] which corroborates our observations of -RgIB-treated hyperactive mice.
In conclusion, we have isolated a novel conotoxin from Conus regius and, by means of a combination of biochemical, structural and pharmacological assays were able to classify this peptide in the -family and named it -RgIB. There are still several peptides to be explored in the C. regius venom, as our previous qualitative investigations have shown [34] and the current study has focused on the biochemical characterization of one such novel peptide. Further studies are still necessary to better characterize the structural and pharmacological properties of -RgIB.